Control method, device, and system for wind turbine set, and storage medium

Abstract

The present disclosure provides a control method, device, and system for a wind turbine set, and a storage medium. The control method comprises: acquiring a preset value of a control parameter and a plurality of historical measurement signals of a wind turbine set in a historical time period; according to the preset value and the plurality of historical measurement signals, determining an estimated distribution parameter of a control signal at a current moment; and determining the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment.

Description

Control method, device, system and storage medium of wind turbine generator set Technical Field

[0001] The present disclosure relates to the field of wind power generation, and more particularly, to a control method, device, system and storage medium for a wind turbine generator set. Background Art

[0002] Current wind turbine control systems mostly use PID control, which employs proportional, integral, and differential control. This linear combination of the proportional, integral, and differential components of the deviation between the preset control parameter and the actual output value forms the control variable, which controls the controlled object. However, the actual output value is currently measured by sensors, which can introduce uncertainty and lead to inaccurate control. Replacing high-precision sensors is often costly.

[0003] Summary of the Invention

[0004] Therefore, it is crucial to improve the control accuracy of wind turbines at low cost.

[0005] In a general aspect, a control method for a wind turbine generator set is provided, comprising: obtaining preset values ​​of control parameters of the wind turbine generator set within a historical period and a plurality of historical measurement signals; determining estimated distribution parameters of the control signal at a current moment based on the preset values ​​and the plurality of historical measurement signals; and determining the control signal at a current moment based on the estimated distribution parameters of the control signal at a current moment.

[0006] In another general aspect, a control device for a wind turbine generator set is provided, comprising: an acquisition unit configured to acquire preset values ​​of control parameters of the wind turbine generator set within a historical period and a plurality of historical measurement signals; a determination unit configured to determine an estimated distribution parameter of the control signal at a current moment based on the preset values ​​and the plurality of historical measurement signals; and a control unit configured to determine the control signal at a current moment based on the estimated distribution parameter of the control signal at a current moment.

[0007] In another general aspect, a control system for a wind turbine generator set is provided, comprising: a measurement sensor for collecting measurement signals of control parameters; a processor for obtaining preset values ​​of the control parameters of the wind turbine generator set within a historical period and a plurality of historical measurement signals; determining an estimated distribution parameter of the control signal at a current moment based on the preset values ​​and the plurality of historical measurement signals; and a control decision maker for determining the control signal at a current moment based on the estimated distribution parameter of the control signal at a current moment.

[0008] In another general aspect, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to perform the control method of the wind turbine generator set as described above.

[0009] In another general aspect, a computer device is provided, comprising: at least one processor; and at least one memory storing computer-executable instructions, wherein the computer-executable instructions, when executed by the at least one processor, cause the at least one processor to execute the control method for a wind turbine generator set as described above.

[0010] The present disclosure proposes a control method, device, system and storage medium for a wind turbine generator set. By obtaining preset values ​​of control parameters and multiple historical measurement signals within a historical period, the uncertainty of the measurement signal at the current moment can be reflected with the help of multiple historical measurement signals, and then the estimated distribution parameters of the control signal at the current moment are obtained in combination with the preset values ​​of the control parameters. Compared with the traditional solution of directly determining the control signal at the current moment based on the measurement signal at the current moment, the control signal can be given from a probabilistic and statistical perspective while fully considering the uncertainty of the measurement signal, thereby reducing the influence of the uncertainty of traditional measurement sensors on the control accuracy, eliminating the need to replace expensive high-precision sensors, and achieving low-cost control accuracy improvement.

[0011] In addition, by obtaining the distribution parameters of the control deviation within the reference period, it is possible to predict the possible control deviation of the new control signal in advance from a probabilistic statistical perspective, and then make corresponding corrections to the estimated distribution parameters of the control signal at the current moment, which can further improve the control accuracy.

[0012] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] FIG1 is a schematic diagram showing a control flow executed by a control system of a wind turbine generator set according to a specific embodiment of the present disclosure;

[0014] FIG2 is a flow chart illustrating a method for controlling a wind turbine generator system according to an embodiment of the present disclosure;

[0015] FIG3 is a schematic diagram showing a probability density function of a measurement signal according to an embodiment of the present disclosure;

[0016] FIG4 is a schematic flow chart illustrating control deviation analysis according to an embodiment of the present disclosure;

[0017] FIG5 is a schematic diagram showing control deviation statistics according to an embodiment of the present disclosure;

[0018] FIG6 is a schematic diagram showing a control deviation distribution according to an embodiment of the present disclosure;

[0019] FIG7 is a block diagram showing a control device of a wind turbine generator set according to an embodiment of the present disclosure;

[0020] FIG8 is a block diagram illustrating a computer device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] The following detailed description is provided to help the reader gain a comprehensive understanding of the methods, devices and / or systems described herein. However, various changes, modifications and equivalents of the methods, devices and / or systems described herein will be clear after understanding the disclosure of the present application. For example, the order of operations described herein is merely an example and is not limited to those orders set forth herein, but can be changed as will be clear after understanding the disclosure of the present application, except for operations that must occur in a specific order. In addition, for greater clarity and conciseness, descriptions of features known in the art may be omitted.

[0022] The features described herein can be implemented in different forms and should not be construed as limited to the examples described herein. Rather, the examples described herein are provided to illustrate only some of the many possible ways to implement the methods, devices, and / or systems described herein, which will become clear after understanding the disclosure of this application.

[0023] As used herein, the term "and / or" includes any one of the associated listed items and any combination of any two or more.

[0024] Although terms such as "first," "second," and "third" may be used herein to describe various members, components, regions, layers, or portions, these members, components, regions, layers, or portions should not be limited by these terms. Instead, these terms are used solely to distinguish one member, component, region, layer, or portion from another member, component, region, layer, or portion. Thus, what is referred to as a first member, first component, first region, first layer, or first portion in the examples described herein may also be referred to as a second member, second component, second region, second layer, or second portion without departing from the teachings of the examples.

[0025] In the specification, when an element (such as a layer, region, or substrate) is described as being “on,” “connected to,” or “coupled to” another element, the element may be directly “on,” “connected to,” or “coupled to” the other element, or one or more other elements may be present therebetween. Conversely, when an element is described as being “directly on,” “directly connected to,” or “directly coupled to” another element, there may be no other elements present therebetween.

[0026] The terms used herein are only used to describe various examples and are not intended to limit the disclosure. Unless the context clearly indicates otherwise, the singular is intended to include the plural. The terms "comprise," "include," and "have" indicate the presence of the recited features, quantities, operations, components, elements, and / or combinations thereof, but do not preclude the presence or addition of one or more other features, quantities, operations, components, elements, and / or combinations thereof.

[0027] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure pertains after understanding the present disclosure. Unless expressly defined otherwise herein, terms (such as those defined in general dictionaries) should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and should not be interpreted in an idealized or overly formal manner.

[0028] Furthermore, in describing the examples, when it is deemed that a detailed description of well-known related structures or functions would cause ambiguous interpretation of the present disclosure, such detailed description will be omitted.

[0029] The present disclosure proposes a control system for a wind turbine generator set, including a measuring sensor, a processor and a control decision maker. The measuring sensor is used to collect measurement signals of control parameters, and a traditional measuring sensor can be used to reduce costs. The processor is used to obtain preset values ​​of control parameters of the wind turbine generator set in a historical period and multiple historical measurement signals, thereby reflecting the uncertainty of the measurement signal, and determining the estimated distribution parameters of the control signal at the current moment based on the preset values ​​and multiple historical measurement signals. The traditional PID control algorithm can be applied to realize the determination of the control signal based on the preset value and the measurement signal. On this basis, the statistical algorithm can also be used to process the determined multiple control signals to obtain the estimated distribution parameters of the control signal at the current moment. The control decision maker is used to determine the control signal at the current moment based on the estimated distribution parameters of the control signal at the current moment, realize the provision of the control signal from the perspective of probability statistics, and further control the operation of the wind turbine generator set based on the control signal.

[0030] Compared with the traditional control system that directly determines the control signal at the current moment based on the measurement signal at the current moment, the control system proposed in the present disclosure can give the control signal from a probabilistic statistical perspective while fully considering the uncertainty of the measurement signal, thereby reducing the impact of the uncertainty of the traditional measurement sensor on the control accuracy. There is no need to replace expensive high-precision sensors, thereby achieving low-cost improvement in control accuracy.

[0031] It should be understood that the control system proposed in the present disclosure can be a stand-alone control system for a wind turbine generator set or a field-level control system for a wind farm. The measurement sensor is a separate sensor device, and the processor and control decision maker are both used for data processing. They can be two separate chips and connected to each other in a wired or wireless manner (hereinafter simply described as connected, unless otherwise specified, both indicate a wired or wireless connection) to achieve data transmission; the two can also be integrated into the same chip. In addition, it is not difficult to find that the processor can further include a first signal analyzer and a first controller, and can further include a second signal analyzer and a second controller. In this regard, the components contained in the processor can be split and made into at least two chips, for example, the first signal analyzer, the first controller, the second signal analyzer, and the second controller can be made into four independent and interconnected chips, or the first signal analyzer and the second signal analyzer can be integrated into a signal analyzer chip, and the first controller and the second controller can be integrated into a controller chip; further, the control decision maker can also be integrated into a chip with the first controller and / or the second controller. The above is an exemplary description of the implementation of each component in the control system of the present disclosure, and is not a limitation of the present disclosure.

[0032] Optionally, the processor is further configured to: obtain a reference measurement signal based on a plurality of historical measurement signals; and determine an estimated distribution parameter of the control signal at a current moment based on each reference measurement signal and a preset value.

[0033] Optionally, the processor includes: a first signal analyzer, used to perform fitting processing on multiple historical measurement signals to obtain distribution parameters of the historical measurement signals; a first controller, used to discretize the distribution parameters of the historical measurement signals to obtain multiple reference measurement signals and a reference measurement signal probability of each reference measurement signal.

[0034] Optionally, the first controller is also used to: determine the reference control signal corresponding to the reference measurement signal based on each reference measurement signal and a preset value, and use the reference measurement signal probability of the reference measurement signal as the reference control signal probability; perform fitting processing on the determined multiple reference control signals and the reference control signal probability of each reference control signal to obtain the estimated distribution parameters of the control signal at the current moment.

[0035] Optionally, the first signal analyzer is also used to: perform fitting processing on multiple historical measurement signals according to a reference distribution type to obtain distribution parameters of the historical measurement signals, wherein the reference distribution type is related to the measurement sensor; or the reference distribution type is obtained by the following steps: performing frequency statistics on multiple historical measurement signals; and determining one from multiple candidate distribution types based on the frequency statistical results of the multiple historical measurement signals as the reference distribution type.

[0036] Optionally, the processor also includes: a second signal analyzer, used to obtain the distribution parameters of the control deviation within a reference time period, wherein the control deviation is the deviation between the measured signal and the preset value; a second controller, used to correct the estimated distribution parameters of the control signal at the current moment based on the distribution parameters of the control deviation within the reference time period, and obtain the corrected distribution parameters of the control signal at the current moment; the control decision maker is also used to determine the control signal at the current moment based on the corrected distribution parameters of the control signal at the current moment.

[0037] Optionally, the second signal analyzer is also used to: determine a number of consecutively arranged preset time lengths from the starting moment; perform frequency statistics on the control deviations within the first i preset time lengths to obtain the i-th statistical result; i is a positive integer; perform fitting processing on the i-th statistical result to obtain the distribution parameter of the i-th control deviation; in response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation not satisfying the preset approximate conditions, increase i by 1, and repeat the steps of performing frequency statistics, fitting processing, and judging the preset approximate conditions on the control deviations within the first i preset time lengths; in response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation satisfying the preset approximate conditions, use the distribution parameter of the i-th control deviation as the distribution parameter of the control deviation within the reference period, use the period corresponding to the first i preset time lengths as the reference period, and use the end time of the i-th preset time length as the starting time of the new reference period, and for the new reference period, repeat the steps of performing frequency statistics, fitting processing, and judging the preset approximate conditions on the control deviations within the first i preset time lengths.

[0038] Optionally, the second controller is also used to: discretize the estimated distribution parameters of the control signal at the current moment to obtain multiple control signals and the control signal probability of each control signal; discretize the distribution parameters of the control deviation in the reference time period to obtain multiple control deviations and the control deviation probability of each control deviation; perform superposition correction processing on each control signal and each control deviation to obtain a corresponding corrected control signal, and determine the product of the control signal probability of the control signal and the control deviation probability of the control deviation as the correction probability of the corresponding corrected control signal; perform fitting processing on the obtained multiple corrected control signals and the correction probability of each corrected control signal to obtain the corrected distribution parameters of the control signal at the current moment.

[0039] Optionally, the control decision maker is further configured to determine the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range.

[0040] Optionally, the control decision maker is also used to: determine the probability that the load falls within the reference load range as a reference probability based on the estimated distribution parameters at the current moment, the correspondence between the control signal and the load, and the reference load range; determine the reference control signal range corresponding to the reference load range in response to the reference probability being greater than or equal to the probability threshold; determine the power generation corresponding to multiple reference control signals within the reference control signal range based on the correspondence between the control signal and the power generation; and determine the control signal at the current moment from multiple reference control signals based on the power generation corresponding to the multiple reference control signals.

[0041] Optionally, the control decision maker is further configured to determine, in response to the reference probability being less than the probability threshold, a reference control signal corresponding to an upper limit value of the reference load range as the control signal at the current moment.

[0042] Figure 1 is a schematic diagram illustrating the control flow executed by a control system for a wind turbine generator system according to a specific embodiment of the present disclosure. In this embodiment, the control system includes conventional measurement sensors, a processor, and a control decision maker. The processor includes a first signal analyzer, a first controller, a second signal analyzer, and a second controller. The first signal analyzer and the second signal analyzer are integrated into a single signal analyzer chip, and the first controller and the second controller are integrated into a single controller chip.

[0043] Referring to Figure 1, the control system is run for any of the control parameters, taking pitch angle, rotational speed, torque, and yaw angle as control parameters. The reasons for inaccurate control can be divided into two categories: one is the uncertainty in the measurement process, and the other is the control deviation between the preset value of the control parameter and the measurement signal detected after the actual unit is executed. Accordingly, on the one hand, the preset value of the control parameter in the historical period and multiple historical measurement signals are input into the first signal analyzer, and the first signal analyzer performs measurement uncertainty analysis to obtain the distribution parameters of the historical measurement signals; the first controller obtains the feedback data of the PID control algorithm based on this, so that the measurement uncertainty that may be brought by the measurement sensor can be taken into account when running the PID control algorithm, and then the control signal is output, and the estimated distribution parameters of the control signal at the current moment are statistically obtained, which compensates for the influence of measurement uncertainty. On the other hand, starting from the starting moment of the reference time period, the accumulated control deviation is input into the second signal analyzer at every preset time period, and the second signal analyzer performs control deviation analysis to obtain the distribution parameters of the control deviation in the reference time period. Based on this, the estimated distribution parameters of the control signal obtained by the first controller can be subjected to lag correction to obtain the corrected distribution parameters of the control signal, which not only retains the robustness of traditional PID control but also achieves the effect of predictive control.

[0044] On this basis, the control decision maker combines the existing SCADA TOLOAD method, which derives the unit load from measurement signals such as the unit's pitch angle, speed, torque, and yaw angle. This method can obtain the load distribution parameters corresponding to the modified distribution parameters of the control signal. Given a reference load range for a critical load component, the reference probability that the critical load component falls within the reference load range can be calculated based on the load distribution parameters. Combined with the reference probability, if the reference probability is greater than or equal to the probability threshold, the control solution is optimized within the reference control signal range corresponding to the reference load range, using power generation as a reference. A control solution with safe load and relatively optimal power generation is determined, maximizing power generation while ensuring safe operation. If the reference probability is less than the probability threshold, the control signal that causes the load at the upper limit of the reference load range is directly inferred based on the upper limit of the reference load range to determine the final control solution. This allows for rapid determination of the control solution and implementation of unit control, making control more accurate and timely.

[0045] FIG2 is a flow chart illustrating a method for controlling a wind turbine generator system according to an embodiment of the present disclosure.

[0046] 2 , in step S101 , preset values ​​of control parameters of a wind turbine generator set within a historical period and a plurality of historical measurement signals are acquired.

[0047] Control parameters include, but are not limited to, pitch angle, rotational speed, torque, and yaw angle. A preset value is the target value of the control parameter that is desired to be achieved through control. For example, if the pitch angle is desired to be adjusted to 0°, the preset value of the pitch angle is 0°. This preset value often remains unchanged for a period of time. As the name suggests, a historical measurement signal is a measurement signal over a historical period of time. A measurement signal is an electrical signal corresponding to a control parameter and can be obtained by measuring the control parameter using a measurement sensor. During the control process, existing control methods can output a control signal based on the preset value of the control parameter. This control signal, when applied to the corresponding actuator, can cause the corresponding control parameter of the actuator to change. By using a measurement sensor, the actual value of the corresponding control parameter can be measured as a measurement signal. It can be seen that at each moment, the preset value of the control parameter, the control signal, and the measurement signal correspond to each other, and both the control signal and the measurement signal fluctuate around the preset value. When the preset value changes, the control signal and the measurement signal also change accordingly.

[0048] The present disclosure obtains multiple historical measurement signals, and can replace the measurement signal at the current moment with multiple historical measurement signals when controlling at the current moment, thereby reflecting the uncertainty of the measurement signal at the current moment. It should be understood that in order to ensure that the multiple historical measurement signals obtained can reflect the uncertainty of the measurement signal at the current moment, it is necessary to at least keep the preset values ​​corresponding to these multiple historical measurement signals unchanged and consistent with the preset values ​​at the current moment. Preferably, a historical period close to the current moment can be selected, such as a historical period 10 minutes before the current moment. Since the historical measurement signal is close to the current moment, the possibility of being affected by other factors is relatively small, and it can better reflect the measurement uncertainty of the current moment.

[0049] Step S102 : determining an estimated distribution parameter of the control signal at the current moment according to a preset value and a plurality of historical measurement signals.

[0050] This step takes into account the uncertainty of the measurement signal at the current moment when performing control. Accordingly, what is determined is the estimated distribution parameters of the control signal at the current moment, rather than a single clear control signal, which can achieve probabilistic control. As an example, multiple historical measurement signals can be directly used as input data of the traditional control algorithm, such as as feedback data in the PID algorithm. According to each historical measurement signal and the preset value, the control signal corresponding to each historical measurement signal is determined, and then the multiple control signals determined are statistically analyzed to obtain the estimated distribution parameters of the control signal at the current moment. This processing method not only continues the traditional control algorithm, but also can use multiple historical measurement signals to reflect the uncertainty of the measurement signal, thereby fully considering the uncertainty of the measurement signal and providing a control signal from the perspective of probabilistic statistics. It can reduce the impact of the uncertainty of the traditional measurement sensor on the control accuracy, without the need to replace expensive high-precision sensors, and achieve low-cost, high-precision control.

[0051] As an example, the estimated distribution parameters are parameters in the probability density function (PDF), such as the location parameter μ and scale parameter σ in the normal distribution. Of course, other parameters that can reflect the characteristics of the distribution can also be used, and this disclosure is not limited to this. The various distribution parameters of other data are similarly described and will not be described one by one below.

[0052] Step S103: Determine the control signal at the current moment based on the estimated distribution parameters of the control signal at the current moment. By combining the estimated distribution parameters of the control signal at the current moment, a control signal can be given from a probability and statistical perspective, thereby improving the accuracy of the obtained control signal.

[0053] Next, step S102 will be further introduced.

[0054] Optionally, step S102 includes: obtaining a reference measurement signal based on multiple historical measurement signals; and determining an estimated distribution parameter of the control signal at the current moment based on each reference measurement signal and a preset value. By determining multiple reference measurement signals based on the historical measurement signals and using the reference measurement signals as input data for a conventional control algorithm, the multiple historical measurement signals can be further processed to adjust the accuracy of the estimated distribution parameter of the control signal at the current moment.

[0055] Regarding how to obtain a reference measurement signal, in some embodiments, at least a portion of the historical measurement signals may be selected from multiple historical measurement signals to serve as the reference measurement signal. For example, all of the multiple historical measurement signals may be directly used as the reference measurement signal, or a portion of the signals may be selected according to a certain rule. The rule may include, but is not limited to, removing the maximum and minimum values, which is not limited in this disclosure.

[0056] In other embodiments, multiple historical measurement signals are first fitted to obtain distribution parameters of the historical measurement signals; the distribution parameters of the historical measurement signals are then discretized to obtain multiple reference measurement signals and a reference measurement signal probability for each reference measurement signal. By first fitting the distribution parameters of the historical measurement signals, the distribution law followed by the historical measurement signals can be extracted from a limited number of historical measurement signals, achieving the distillation of concrete data into abstract laws. Discretizing the extracted distribution parameters can then ensure that the multiple reference measurement signals obtained conform to the extracted distribution parameters, that is, conform to the abstract laws, thereby improving the accuracy of the representation of measurement uncertainty, helping to improve the accuracy of the estimated distribution parameters of the control signal at the current moment, and thereby improving control accuracy.

[0057] For some other embodiments described above, correspondingly, the step of determining the estimated distribution parameters of the control signal at the current moment includes: determining the reference control signal corresponding to the reference measurement signal based on each reference measurement signal and a preset value, and using the reference measurement signal probability of the reference measurement signal as the reference control signal probability; performing fitting processing on the determined multiple reference control signals and the reference control signal probability of each reference control signal to obtain the estimated distribution parameters of the control signal at the current moment. Based on the distribution parameters of the refined historical measurement signals, the reference measurement signal probability of each reference measurement signal can be obtained. By using the reference measurement signal probability of each reference measurement signal as the reference control signal probability of the reference control signal corresponding to the reference measurement signal, a more accurate probability description of the determined multiple reference control signals can be performed based on the analyzed measurement uncertainty, thereby ensuring the accuracy of the estimated distribution parameters of the control signal at the current moment. It should be understood that since the values ​​of the reference measurement signal and its corresponding reference control signal are often not absolutely equal, the distribution type to which the estimated distribution parameters of the control signal at the current moment belong may be the same as or different from the distribution type to which the distribution parameters of the historical measurement signals belong. The distribution type here refers to the probability distribution type, such as normal distribution and Gaussian distribution.

[0058] In addition, the fitting processing of multiple historical measurement signals in some other embodiments described above can be performed according to the reference distribution type, so that the distribution parameters of the historical measurement signals can be obtained relatively reliably by determining the parameters in the reference distribution type, thereby realizing measurement uncertainty analysis.

[0059] As an example, the reference distribution type is triangular distribution, Measure refers to the measurement signal, and the maximum and minimum values ​​of multiple historical measurement signals [Max Measure 、Min Measure ], according to [Max Measure 、Min Measure ] A symmetric triangular distribution can be constructed, and its probability density function f(Measure) is shown in the following formula, and the schematic diagram is shown in Figure 3.

[0060] Regarding the determination of the reference distribution type, given that multiple historical measurement signals are measured by measurement sensors, and the uncertainty is often caused by the errors in the measurement sensors themselves, the distribution type of the measurement signals is likely to be related to the measurement sensors. Based on this, in one example, the reference distribution type is related to the measurement sensor. By analyzing the distribution types of signals measured by different measurement sensors in advance, and then determining the corresponding distribution type based on the measurement sensor used to detect the measurement signal, a relatively reliable reference distribution type can be quickly and conveniently determined as the reference distribution type, thereby improving data analysis efficiency. As an example, different distribution types can be matched according to the type of measurement sensor, and different distribution types can be further matched for measurement sensors of the same type but different precisions in combination with the accuracy of the measurement sensor. This is not limited by the present disclosure.

[0061] In another example, the reference distribution type is obtained by the following steps: performing frequency statistics on multiple historical measurement signals; based on the frequency statistics results of the multiple historical measurement signals, determining one from multiple candidate distribution types as the reference distribution type. By actually analyzing the distribution of multiple historical measurement signals obtained (i.e., performing frequency statistics), and then determining a similar one from multiple candidate distribution types, it is possible to independently analyze the measurement signal obtained each time, thereby increasing the possibility that the determined reference distribution type conforms to the distribution characteristics of the historical measurement signal. As an example, the cumulative probability distribution (CPD) corresponding to each value can be statistically calculated based on the frequency of each historical measurement signal (i.e., the number of times different historical measurement signal values ​​appear), and then for each candidate distribution type, the cumulative probability distribution value of each measurement signal value is determined, and then a candidate distribution type that is relatively close to the statistical result is selected as the reference distribution type. Alternatively, a distribution diagram of the historical measurement signal can be plotted using the values ​​of the historical measurement signal as the horizontal axis and the frequency of each value as the vertical axis. This can then be compared with the probability density function curves of each candidate distribution type. For example, the degree of overlap between the curve trends can be calculated, and the candidate distribution type with the greatest overlap can be selected as the reference distribution type. This disclosure is not limited to this.

[0062] As mentioned above, the causes of inaccurate control can be categorized into two types: uncertainty in the measurement process, and control deviation between the preset values ​​of the control parameters and the measured signals detected after actual unit execution. Step S102 addresses the issue of measurement uncertainty, and the correction of control deviation is described below.

[0063] With respect to the control deviation, before step S103, the control method of an embodiment of the present disclosure may further include: obtaining the distribution parameters of the control deviation within a reference time period, wherein the control deviation is the deviation between the measured signal and the preset value; accordingly, step S103 includes: based on the distribution parameters of the control deviation within the reference time period, correcting the estimated distribution parameters of the control signal at the current moment to obtain the corrected distribution parameters of the control signal at the current moment; and determining the control signal at the current moment according to the corrected distribution parameters of the control signal at the current moment.

[0064] Control deviation is the difference between the measured signal and the preset value, and can be considered the difference between the actual control result and the expected result. The larger the individual control deviation, the further the corresponding measured signal is from the preset value, the higher the control lag, and the lower the accuracy. By obtaining the control deviation distribution parameters within a reference period, it is possible to predict the possible lag of the control signal in advance from a probabilistic and statistical perspective, and then correct the lagged control signal. This means correcting the estimated distribution parameters of the control signal at the current moment to obtain the corrected distribution parameters of the control signal at the current moment. This allows predictive control of wind turbines in the face of complex and changing wind conditions, improving control accuracy.

[0065] In terms of time, the current moment is the moment when the preset value of the control parameter is executed, that is, the moment when the control is implemented according to the preset value and the lag correction is implemented according to the distribution parameters of the control deviation in the reference period. The end time of the reference period should be earlier than or equal to the current moment, that is, the distribution parameters of the control deviation in the reference period are analyzed and obtained at the earliest at the end time of the reference period, and immediately applied to the lag correction at the current moment, and then the distribution parameters can continue to be applied to the lag correction at subsequent moments; of course, it is also possible to analyze the distribution parameters of the control deviation in the reference period after a period of time, or to apply the analysis results to the current lag correction after a period of time, so that the end time of the reference period is always earlier than the current moment, and the present disclosure does not impose any restrictions on this.

[0066] Optionally, referring to FIG4 , the step of analyzing the control deviation, i.e., the step of obtaining the distribution parameters of the control deviation within the reference period, further includes the following steps:

[0067] The first step is to determine a number of consecutive preset time periods starting from the starting time. It should be understood that a certain amount of control deviation data must be accumulated before analysis can be performed. The starting time is the moment when control deviation data recording begins. By determining a number of consecutive preset time periods, a statistical analysis of the existing control deviation data can be performed in subsequent steps every time control deviation data accumulates for a preset time period. The preset time period is, for example, but not limited to, 10 minutes. It should be noted that this step emphasizes the confirmation of the starting time, which is not specifically shown in Figure 4.

[0068] The second step is to perform frequency statistics on the control deviations within the first i preset time periods, obtaining the i-th statistical result; i is a positive integer. This step performs frequency statistics on all control deviation data accumulated from the start time to the current time every preset time period. Therefore, i will gradually increase from 1 to achieve periodic control deviation data statistics. It should be understood that for each statistical period, the time period consisting of the first i preset time periods is the continuous time period shown in Figure 4.

[0069] As an example, for the first i consecutive time periods of a preset length, the measurement signals of pitch angle, speed, torque and yaw angle [PitchAngle measure 、RotorSpeed measure 、Torque measure 、YawAngle measure ] and the corresponding preset value [PitchAngle design 、RotorSpeed design 、Torque design 、YawAngle design ] is used as input to calculate the control deviation between the measured signal and the preset value. The actual control parameter used can be selected according to the actual control requirements of the unit. As shown in the following formula, Error refers to the control deviation, Measure refers to the measured signal, and Design refers to the control set value. Error = Measure-Design

[0070] Assuming that the control deviation of a certain control parameter in the continuous time period is as shown in FIG5 , the various values ​​of the control deviation in the continuous time period and the frequency corresponding to each value are counted. As the current latest statistical result, a control deviation distribution diagram as shown in FIG6 can also be obtained.

[0071] The third step is to perform fitting on the i-th statistical result to obtain the distribution parameters of the i-th control deviation. The specific fitting method can be found in the previous section on measurement uncertainty analysis. First, a distribution type is determined, and then fitting is performed according to that distribution type to determine the distribution parameters. This will not be repeated here.

[0072] The fourth step is to determine whether the distribution parameters of the i-th control deviation and the distribution parameters of the i-1-th control deviation meet the preset approximation condition. It should be understood that the preset approximation condition represents a condition that the two compared distribution parameters are sufficiently close, for example, but not limited to, the difference between the two compared distribution parameters is less than 5%. For cases involving multiple specific parameters, such as the parameters of the normal distribution including the location parameter μ and the scale parameter σ, one data can be obtained by processing, for example, converting multiple specific parameters into one parameter, or calculating the difference percentage of each specific parameter separately, and then calculating the statistical value of the difference percentage of each specific parameter (such as the mean, mode, median, etc.). The present disclosure does not limit this.

[0073] After the above judgment, if the result does not meet the preset approximate conditions, it can be considered that the currently accumulated data volume is insufficient and does not reflect a consistent distribution pattern. In this case, i is increased by 1, and steps 2 to 4 are repeated. In other words, the periodic control deviation data statistics are continued, and the new control deviation distribution parameters are fitted, and the judgment of whether the preset approximate conditions are met is again made. If the result meets the preset approximate conditions, it can be considered that a stable and consistent distribution pattern has been statistically analyzed. The distribution parameters of the i-th control deviation are used as the distribution parameters of the control deviation within the reference period, and the period corresponding to the first i preset time lengths is used as the reference period. At this time, a round of control deviation analysis is completed. At the same time, the end time of the i-th preset time length is used as the starting time of the new reference period. For the new reference period, steps 1 to 4 are repeated to achieve a new round of control deviation analysis.

[0074] It should be understood that in the fourth step, if i=1, the distribution parameters of the (i-1)th control deviation can use the distribution parameters of the control deviation in the reference period obtained in the previous round of control deviation analysis.

[0075] It should also be understood that after completing a round of control deviation analysis, the obtained distribution parameters of the control deviation within the reference period can be applied to hysteresis correction. At this point, data can be re-accumulated simultaneously to conduct a new round of control deviation analysis until the distribution parameters of the control deviation within the new reference period are obtained. These distribution parameters are then used to perform hysteresis correction. This cycle repeats, with the latest distribution parameters of the control deviation within the reference period continuously used to perform hysteresis correction. Precisely because the obtained distribution parameters of the control deviation within the reference period are used to perform control hysteresis correction, theoretically, the control deviation will show a decreasing trend over time rather than remaining constant over time. By re-accumulating data in a new round of analysis, it is possible to avoid the situation in which the control deviation data from the previous round dominates and quickly meets the preset approximation conditions in the early stages of the new round of analysis. This facilitates effective statistical analysis of the new round of control deviation data, thereby reducing the impact of long-standing large deviation data on recent control deviation analysis results. This fully reflects the recent control optimization results, provides a more reliable basis for updating the control deviation analysis results, and thus improves control accuracy, forming a virtuous cycle.

[0076] Optionally, the step of applying the result of the control deviation analysis to implement the lag correction, that is, based on the distribution parameters of the control deviation in the reference time period, correcting the estimated distribution parameters of the control signal at the current moment to obtain the corrected distribution parameters of the control signal at the current moment, further includes: discretizing the estimated distribution parameters of the control signal at the current moment to obtain multiple control signals and the control signal probability of each control signal; discretizing the distribution parameters of the control deviation in the reference time period to obtain multiple control deviations and the control deviation probability of each control deviation; performing superposition correction processing on each control signal and each control deviation to obtain a corresponding corrected control signal, and determining the product of the control signal probability of the control signal and the control deviation probability of the control deviation as the corrected probability of the corresponding corrected control signal; fitting the obtained multiple corrected control signals and the correction probability of each corrected control signal to obtain the corrected distribution parameters of the control signal at the current moment. The estimated distribution parameters of the control signal at the current moment and the distribution parameters of the control deviation within the reference period are both parameters of continuous functions. By discretizing the two, multiple control signals and their probabilities, as well as multiple control deviations and their probabilities, can be obtained. This facilitates pairing each control signal and each control deviation for superposition correction processing. A single control deviation can be used to compensate for the lag of a single control signal, ensuring the smooth progress of the correction process. It should be understood that, assuming that M control signals and N control deviations are obtained after discretization, M×N data pairs can be obtained, and correspondingly M×N corrected control signals and corrected probabilities can be obtained. These corrected control signals may have equal values. In this case, the corrected control signals with equal values ​​can be merged, and the sum of the corrected probabilities of the corrected control signals with equal values ​​is used as the merged corrected probability.

[0077] As an example, in step S102, multiple historical measurement signals are first fitted and then discretized to obtain multiple reference measurement signals. The estimated distribution parameters of the control signal at the current moment are originally obtained by fitting multiple reference control signals corresponding to the multiple reference measurement signals. In this case, the multiple reference control signals can be fitted without first being fitted in step S102. These multiple reference control signals can be directly used when performing the hysteresis correction in step S103, which can reduce the amount of data processing and improve computational efficiency. It should be understood that although the estimated distribution parameters of the control signal at the current moment are not clearly obtained at this time, the estimated distribution parameters essentially exist, so it still complies with step S102 in Figure 2. Of course, it is also possible to first clearly obtain the estimated distribution parameters of the control signal at the current moment in step S102, and then perform discretization during the hysteresis correction, so as to obtain multiple control signals according to the expected rules, for example, to obtain multiple control signals in an arithmetic progression to meet the different requirements of the superposition correction processing. The present disclosure is not limited to this.

[0078] Next, how to determine the control signal at the current moment in step S103 will be further introduced.

[0079] It should be understood that when the aforementioned hysteresis correction is not performed, the estimated distribution parameters of the current control signal can be directly used to determine the current control signal. When the aforementioned hysteresis correction is performed, the corrected distribution parameters of the current control signal can be used to determine the current control signal. For ease of explanation, the following text uniformly uses the estimated distribution parameters of the current control signal. When hysteresis correction is performed, the estimated distribution parameters of the current control signal are replaced with the corrected distribution parameters of the current control signal, and no further explanation is given.

[0080] Optionally, the step of determining the control signal at the current moment includes: determining the control signal at the current moment based on the estimated distribution parameters of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range. Since specific control parameters are often related to specific load components, changes in the control parameters will cause changes in the corresponding load components. By introducing the reference load range and using the corresponding relationship between the control signal and the load, a reasonable control signal can be determined based on the estimated distribution parameters of the control signal at the current moment with the reference load range as a reference. This can not only ensure that the determined control signal complies with the hysteresis correction, which helps to improve the accuracy of the control, but also ensure that the load component falls within the reference load range after the corresponding control is executed, thereby ensuring the safe operation of the unit and fully improving the control efficiency. It should be understood that to ensure safe operation, the reference load range must have an upper limit value greater than 0, and a lower limit value greater than or equal to 0. The lower limit value can be, for example, given by calculation, or manually entered by the operator, or it can be defaulted to 0. This is not limited in this disclosure.

[0081] Regarding the specific execution of this step, in some embodiments, optionally, a reference control signal range corresponding to the reference load range is first determined based on the correspondence between the control signal and the load; and then, based on the estimated distribution parameters of the control signal at the current moment, a value is selected from the reference control signal range, for example, a value with the largest probability density value, as the control signal at the current moment.

[0082] In other embodiments, the probability of the load falling within the reference load range is optionally determined based on the estimated distribution parameters at the current moment, the correspondence between the control signal and the load, and the reference load range, serving as a reference probability. The reference probability is then compared with a probability threshold, and in response to the reference probability being greater than or equal to the probability threshold, a reference control signal range corresponding to the reference load range is determined. Based on the correspondence between the control signal and the power generation, the power generation corresponding to multiple reference control signals within the reference control signal range is determined. Based on the power generation corresponding to the multiple reference control signals, the control signal at the current moment is determined from the multiple reference control signals. By determining the reference probability of the load falling within the reference load range, the probability of safe operation of the unit can be determined. A larger reference probability indicates a higher probability of safe operation of the unit. For example, a calculated reference probability of 0.9 indicates that 90% of the load falls within the reference load range. In this case, the adjustable range of the control signal, i.e., the reference control signal range, is also larger, providing more flexible selection. Since the load is derived from the control signal, the reference control signal range corresponding to the reference load range can be inferred based on the corresponding relationship between the control signal and the load. Within this range, the control scheme is optimized with the power generation as a reference, and the reference control signal that maximizes the power generation of the unit and does not exceed the reference load range is searched as the optimal control scheme. This embodiment can calmly face the changing and complex wind conditions with high control accuracy while ensuring the safety of the wind turbine generator set, and achieve more power output, thereby improving economic benefits and helping to improve control efficiency in all aspects. It should be understood that the corresponding relationship between the control signal and the power generation can be obtained through simulation of existing methods, which is a mature technology in this field and is not limited by this disclosure. As an example, the corresponding relationship between the control signal and the load can be reflected in the existing SCADATOLOAD method, that is, when it is necessary to derive the load based on the control signal, the SCADATOLOAD method can be applied to process the control signal to obtain the load.

[0083] In some other embodiments described above, further optionally, after completing the comparison between the reference probability and the probability threshold, in response to the reference probability being less than the probability threshold, a reference control signal corresponding to the upper limit of the reference load range is determined as the control signal at the current moment. When the reference probability is small, it can be considered that the load is likely to exceed the reference load range, and there is a high risk. At this time, the adjustable range of the control signal is small. By directly using the reference control signal corresponding to the upper limit of the reference load range as the final control signal, there is no need to calculate the reference control signal range, nor to optimize the control scheme based on the power generation. This can reasonably simplify the control signal determination process, reduce the amount of calculation, improve control efficiency, and reduce the risk of untimely control and dangerous load overload caused by excessive calculation time, making the control more accurate and timely.

[0084] Regarding the determination of the reference probability, the distribution parameters of the load can be determined based on the estimated distribution parameters of the control signal at the current moment and the corresponding relationship between the control signal and the load, and then the probability of the load falling into the reference load range can be determined based on the distribution parameters of the load and the reference load range. For example, the cumulative probability distribution value corresponding to the reference load range can be calculated by the probability density function of the load as the reference probability. Alternatively, the reference control signal range corresponding to the reference load range can be determined based on the corresponding relationship between the control signal and the load, and then the probability of the control signal falling into the reference control signal range can be determined based on the estimated distribution parameters of the control signal at the current moment. For example, the cumulative probability distribution value corresponding to the reference control signal range can be calculated by the probability density function of the control signal as the reference probability. In other words, the reference probability can be calculated directly for the load, or it can be calculated directly for the control signal. Since there is a corresponding relationship between the control signal and the load, the two are theoretically equal, and the present disclosure does not impose any restrictions on this.

[0085] FIG7 is a block diagram illustrating a control device of a wind turbine generator system according to an embodiment of the present disclosure.

[0086] 7 , a control device 700 for a wind turbine generator set includes an acquisition unit 701 , a determination unit 702 , and a control unit 703 .

[0087] The acquisition unit 701 may acquire preset values ​​of control parameters of the wind turbine generator set within a historical period and a plurality of historical measurement signals.

[0088] The determination unit 702 may determine the estimated distribution parameter of the control signal at the current moment according to a preset value and a plurality of historical measurement signals.

[0089] The control unit 703 may determine the control signal at the current moment according to the estimated distribution parameters of the control signal at the current moment.

[0090] Optionally, the determining unit 702 may further: obtain a reference measurement signal according to a plurality of historical measurement signals; and determine an estimated distribution parameter of the control signal at the current moment according to each reference measurement signal and a preset value.

[0091] Optionally, the determining unit 702 may further: perform fitting processing on multiple historical measurement signals to obtain distribution parameters of the historical measurement signals; and discretize the distribution parameters of the historical measurement signals to obtain multiple reference measurement signals and a reference measurement signal probability of each reference measurement signal.

[0092] Optionally, the determination unit 702 may also: determine the reference control signal corresponding to the reference measurement signal based on each reference measurement signal and a preset value, and use the reference measurement signal probability of the reference measurement signal as the reference control signal probability; perform fitting processing on the determined multiple reference control signals and the reference control signal probability of each reference control signal to obtain the estimated distribution parameters of the control signal at the current moment.

[0093] Optionally, the multiple historical measurement signals are obtained by measuring a measurement sensor, and the determination unit 702 may further: perform fitting processing on the multiple historical measurement signals according to a reference distribution type to obtain distribution parameters of the historical measurement signals, wherein the reference distribution type is related to the measurement sensor; or the reference distribution type is obtained by the following steps: performing frequency statistics on the multiple historical measurement signals; and determining one from multiple candidate distribution types based on the frequency statistics results of the multiple historical measurement signals as the reference distribution type.

[0094] Optionally, the acquisition unit 701 can also: obtain the distribution parameters of the control deviation within the reference time period, wherein the control deviation is the deviation between the measured signal and the preset value; the control unit 703 can also: based on the distribution parameters of the control deviation within the reference time period, correct the estimated distribution parameters of the control signal at the current moment to obtain the corrected distribution parameters of the control signal at the current moment; determine the control signal at the current moment based on the corrected distribution parameters of the control signal at the current moment.

[0095] Optionally, the acquisition unit 701 may also: determine a number of consecutively arranged preset time lengths starting from the starting moment; perform frequency statistics on the control deviations within the first i preset time lengths to obtain the i-th statistical result; i is a positive integer; perform fitting processing on the i-th statistical result to obtain the distribution parameter of the i-th control deviation; in response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation not satisfying the preset approximate condition, increase i by 1, and repeat the steps of performing frequency statistics, fitting processing, and judging the preset approximate condition on the control deviations within the first i preset time lengths; in response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation satisfying the preset approximate condition, use the distribution parameter of the i-th control deviation as the distribution parameter of the control deviation within the reference period, use the period corresponding to the first i preset time lengths as the reference period, and use the end time of the i-th preset time length as the starting time of the new reference period, and repeat the steps of performing frequency statistics, fitting processing, and judging the preset approximate condition on the control deviations within the first i preset time lengths for the new reference period.

[0096] Optionally, the control unit 703 may also: discretize the estimated distribution parameters of the control signal at the current moment to obtain multiple control signals and the control signal probability of each control signal; discretize the distribution parameters of the control deviation within the reference time period to obtain multiple control deviations and the control deviation probability of each control deviation; perform superposition correction processing on each control signal and each control deviation to obtain a corresponding corrected control signal, and determine the product of the control signal probability of the control signal and the control deviation probability of the control deviation as the correction probability of the corresponding corrected control signal; perform fitting processing on the obtained multiple corrected control signals and the correction probability of each corrected control signal to obtain the corrected distribution parameters of the control signal at the current moment.

[0097] Optionally, the control unit 703 may further determine the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range.

[0098] Optionally, the control unit 703 may also: determine the probability that the load falls within the reference load range as a reference probability based on the estimated distribution parameters at the current moment, the correspondence between the control signal and the load, and the reference load range; determine the reference control signal range corresponding to the reference load range in response to the reference probability being greater than or equal to the probability threshold; determine the power generation corresponding to multiple reference control signals within the reference control signal range based on the correspondence between the control signal and the power generation; and determine the control signal at the current moment from multiple reference control signals based on the power generation corresponding to the multiple reference control signals.

[0099] Optionally, the control unit 703 may further: in response to the reference probability being less than the probability threshold, determine a reference control signal corresponding to the upper limit value of the reference load range as the control signal at the current moment.

[0100] Regarding the apparatus in the above embodiment, the specific manner in which each unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated on here.

[0101] The control method of the wind turbine generator set according to the embodiment of the present disclosure can be written as a computer program and stored on a computer-readable storage medium. When the instructions corresponding to the computer program are executed by the processor, the control method of the wind turbine generator set as described above can be implemented. Examples of computer-readable storage media include: read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disk storage, hard disk drive (HDD), solid state drive (SSD), card storage (such as, multimedia card, secure digital (SD) card or ultra fast digital (XD) card), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk and any other device, any other device configured to store the computer program and any associated data, data files and data structures in a non-transitory manner and provide the computer program and any associated data, data files and data structures to a processor or computer so that the processor or computer can execute the computer program. In one example, the computer program and any associated data, data files and data structures are distributed on a networked computer system so that the computer program and any associated data, data files and data structures are stored, accessed and executed in a distributed manner by one or more processors or computers.

[0102] FIG8 is a block diagram illustrating a computer device according to an embodiment of the present disclosure.

[0103] 8 , a computer device 800 includes at least one memory 801 and at least one processor 802 . The at least one memory 801 stores a set of computer-executable instructions. When the computer-executable instruction set is executed by the at least one processor 802 , a control method for a wind turbine generator set according to an exemplary embodiment of the present disclosure is executed.

[0104] As an example, the computer device 800 can be a PC, a tablet device, a personal digital assistant, a smart phone, or other device capable of executing the above-mentioned instruction set. Here, the computer device 800 is not necessarily a single electronic device, but can also be any collection of devices or circuits capable of executing the above-mentioned instructions (or instruction sets) individually or in combination. The computer device 800 can also be part of an integrated control system or system manager, or can be configured as a portable electronic device that is interconnected with a local or remote (e.g., via wireless transmission) interface.

[0105] In computer device 800, processor 802 may include a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller, or a microprocessor. By way of example and not limitation, the processor may also include an analog processor, a digital processor, a microprocessor, a multi-core processor, a processor array, a network processor, etc.

[0106] The processor 802 can execute instructions or codes stored in the memory 801, wherein the memory 801 can also store data. Instructions and data can also be sent and received over the network via the network interface device, wherein the network interface device can use any known transmission protocol.

[0107] It should be understood that the processor in the control system of the wind turbine generator system according to the embodiment of the present disclosure may have a hardware structure similar to the processor 802 herein.

[0108] The memory 801 can be integrated with the processor 802, for example, by placing RAM or flash memory within an integrated circuit microprocessor or the like. Furthermore, the memory 801 can include a separate device, such as an external disk drive, a storage array, or any other storage device usable by a database system. The memory 801 and the processor 802 can be operatively coupled or can communicate with each other, for example, via an I / O port, a network connection, or the like, such that the processor 802 can access files stored in the memory.

[0109] In addition, the computer device 800 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, a mouse, a touch input device, etc.) All components of the computer device 800 may be connected to each other via a bus and / or a network.

[0110] The present disclosure proposes a control method, device, system and storage medium for a wind turbine generator set. By obtaining preset values ​​of control parameters and multiple historical measurement signals within a historical period, the uncertainty of the measurement signal at the current moment can be reflected with the help of multiple historical measurement signals, and then the estimated distribution parameters of the control signal at the current moment can be obtained in combination with the preset values ​​of the control parameters. Compared with the traditional scheme of directly determining the control signal at the current moment based on the measurement signal at the current moment, the control signal can be given from a probability and statistics perspective while fully considering the uncertainty of the measurement signal, thereby reducing the impact of the measurement uncertainty of the traditional measurement sensor on the control accuracy, eliminating the need to replace expensive high-precision sensors, and achieving low-cost control accuracy improvement. In addition, by obtaining the distribution parameters of the control deviation within the reference period, the possible control deviation of the new control signal can be predicted in advance from a probability and statistics perspective, and then the estimated distribution parameters of the control signal at the current moment can be corrected accordingly, which can further improve the control accuracy.

[0111] The specific implementation methods of the present disclosure have been described in detail above. Although some embodiments have been shown and described, those skilled in the art should understand that these embodiments may be modified and varied without departing from the principles and spirit of the present disclosure, the scope of which is defined by the claims and their equivalents. These modifications and variations should also be within the scope of protection of the claims of the present disclosure.

Claims

1. A method for controlling a wind turbine generator set, comprising: Obtaining preset values ​​of control parameters of the wind turbine generator set within a historical period and multiple historical measurement signals; Determining an estimated distribution parameter of the control signal at a current moment according to the preset value and the plurality of historical measurement signals; The control signal at the current moment is determined according to the estimated distribution parameter of the control signal at the current moment.

2. The control method according to claim 1, wherein: The determining, based on the preset value and the plurality of historical measurement signals, an estimated distribution parameter of the control signal at the current moment includes: acquiring a reference measurement signal according to the plurality of historical measurement signals; The estimated distribution parameter of the control signal at the current moment is determined according to each reference measurement signal and the preset value.

3. The control method according to claim 2, wherein: The acquiring of a reference measurement signal according to the plurality of historical measurement signals comprises: performing fitting processing on the plurality of historical measurement signals to obtain distribution parameters of the historical measurement signals; Discretization processing is performed on the distribution parameters of the historical measurement signal to obtain a plurality of reference measurement signals and a reference measurement signal probability of each reference measurement signal.

4. The control method according to claim 3, wherein: The determining, according to each reference measurement signal and the preset value, the estimated distribution parameter of the control signal at the current moment includes: Determining, according to each reference measurement signal and the preset value, a reference control signal corresponding to the reference measurement signal, and using a reference measurement signal probability of the reference measurement signal as a reference control signal probability; Fitting processing is performed on the determined multiple reference control signals and the reference control signal probability of each reference control signal to obtain the estimated distribution parameter of the control signal at the current moment.

5. The control method according to claim 3, wherein: The multiple historical measurement signals are obtained by measurement by a measurement sensor, wherein the fitting process is performed on the multiple historical measurement signals to obtain distribution parameters of the historical measurement signals, including: According to the reference distribution type, fitting processing is performed on the multiple historical measurement signals to obtain distribution parameters of the historical measurement signals, wherein the reference distribution type is associated with the measurement sensor; or The reference distribution type is obtained by the following steps: performing frequency statistics on the multiple historical measurement signals; Based on frequency statistics results of the multiple historical measurement signals, one is determined from multiple candidate distribution types as the reference distribution type.

6. The control method according to any one of claims 1 to 5, wherein: Before determining the control signal at the current moment based on the estimated distribution parameter of the control signal at the current moment, the control method further includes: Obtaining a distribution parameter of a control deviation within a reference period, wherein the control deviation is a deviation between a measurement signal and a preset value; The step of determining the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment includes: Based on the distribution parameters of the control deviation in the reference period, the estimated distribution parameters of the control signal at the current moment are corrected to obtain the corrected distribution parameters of the control signal at the current moment; The control signal at the current moment is determined according to the modified distribution parameter of the control signal at the current moment.

7. The control method according to claim 6, wherein: The obtaining of distribution parameters of the control deviation within the reference period includes: Determine a number of preset durations of consecutive arrangements starting from the starting moment; Perform frequency statistics on the control deviations within the first i preset time periods to obtain the i-th statistical result; i is a positive integer; Perform fitting processing on the i-th statistical result to obtain the distribution parameters of the i-th control deviation; In response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation not satisfying the preset approximation condition, i is increased by 1, and the steps of performing frequency statistics, fitting processing, and determining the preset approximation condition on the control deviations within the first i preset time periods are repeated; In response to the distribution parameters of the i-th control deviation and the distribution parameters of the i-1-th control deviation satisfying the preset approximation conditions, the distribution parameters of the i-th control deviation are used as the distribution parameters of the control deviation within the reference time period, the time period corresponding to the first i preset time lengths is used as the reference time period, and the end time of the i-th preset time length is used as the starting time of the new reference time period. For the new reference time period, the steps of performing frequency statistics, fitting processing and judging the preset approximation conditions on the control deviations within the first i preset time lengths are repeated.

8. The control method according to claim 6, wherein: The step of correcting the estimated distribution parameter of the control signal at the current moment based on the distribution parameter of the control deviation within the reference period to obtain the corrected distribution parameter of the control signal at the current moment includes: Discretizing the estimated distribution parameters of the control signal at the current moment to obtain a plurality of control signals and a control signal probability of each control signal; Discretizing the distribution parameters of the control deviations within the reference period to obtain a plurality of control deviations and a control deviation probability for each control deviation; Performing a superposition correction process on each control signal and each control deviation to obtain a corresponding corrected control signal, and determining the product of the control signal probability of the control signal and the control deviation probability of the control deviation as the correction probability of the corresponding corrected control signal; Fitting processing is performed on the obtained multiple corrected control signals and the corrected probability of each corrected control signal to obtain the corrected distribution parameter of the control signal at the current moment.

9. The control method according to any one of claims 1 to 5, wherein: The determining the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment includes: The control signal at the current moment is determined according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range.

10. The control method according to claim 9, wherein: The determining the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range includes: Determining, based on the estimated distribution parameter at the current moment, the correspondence between the control signal and the load, and the reference load range, a probability that the load falls within the reference load range as a reference probability; In response to the reference probability being greater than or equal to a probability threshold, determining a reference control signal range corresponding to the reference load range; Determining the power generation corresponding to a plurality of reference control signals within the reference control signal range according to the corresponding relationship between the control signal and the power generation; The control signal at the current moment is determined from the multiple reference control signals according to the power generation amounts corresponding to the multiple reference control signals.

11. The control method according to claim 10, wherein: The determining of the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range further includes: In response to the reference probability being less than the probability threshold, a reference control signal corresponding to an upper limit value of the reference load range is determined as the control signal at the current moment.

12. A control device for a wind turbine generator set, comprising: an acquisition unit configured to acquire preset values ​​of control parameters of the wind turbine generator set within a historical period and a plurality of historical measurement signals; a determining unit configured to determine an estimated distribution parameter of the control signal at a current moment based on the preset value and the plurality of historical measurement signals; The control unit is configured to determine the control signal at the current moment according to the estimated distribution parameter of the control signal at the current moment.

13. A control system for a wind turbine generator set, comprising: Measuring sensors for collecting measurement signals of control parameters; Processor for: Obtaining preset values ​​of control parameters of the wind turbine generator set within a historical period and multiple historical measurement signals; Determining an estimated distribution parameter of the control signal at a current moment according to the preset value and the plurality of historical measurement signals; The control decision maker is used to determine the control signal at the current moment according to the estimated distribution parameters of the control signal at the current moment.

14. The control system according to claim 13, wherein: The processor is further configured to: acquiring a reference measurement signal according to the plurality of historical measurement signals; The estimated distribution parameter of the control signal at the current moment is determined according to each reference measurement signal and the preset value.

15. The control system of claim 14, wherein: The processor includes: a first signal analyzer, configured to perform fitting processing on the plurality of historical measurement signals to obtain distribution parameters of the historical measurement signals; The first controller is configured to discretize the distribution parameters of the historical measurement signal to obtain a plurality of reference measurement signals and a reference measurement signal probability of each reference measurement signal.

16. The control system of claim 15, wherein: The first controller is further configured to: Determining, according to each reference measurement signal and the preset value, a reference control signal corresponding to the reference measurement signal, and using a reference measurement signal probability of the reference measurement signal as a reference control signal probability; Fitting processing is performed on the determined multiple reference control signals and the reference control signal probability of each reference control signal to obtain the estimated distribution parameter of the control signal at the current moment.

17. The control system of claim 15, wherein: The first signal analyzer is further configured to: According to the reference distribution type, fitting processing is performed on the multiple historical measurement signals to obtain distribution parameters of the historical measurement signals, wherein the reference distribution type is associated with the measurement sensor; or The reference distribution type is obtained by the following steps: performing frequency statistics on the multiple historical measurement signals; Based on frequency statistics results of the multiple historical measurement signals, one is determined from multiple candidate distribution types as the reference distribution type.

18. The control system according to any one of claims 13 to 17, wherein: The processor further includes: a second signal analyzer, configured to obtain a distribution parameter of a control deviation within a reference period, wherein the control deviation is a deviation between a measurement signal and a preset value; a second controller, configured to correct the estimated distribution parameters of the control signal at the current moment based on the distribution parameters of the control deviation within the reference period, to obtain corrected distribution parameters of the control signal at the current moment; The control decision maker is further configured to determine the control signal at the current moment according to the modified distribution parameter of the control signal at the current moment.

19. The control system of claim 18, wherein: The second signal analyzer is further configured to: Determine a number of preset durations of consecutive arrangements starting from the starting moment; Perform frequency statistics on the control deviations within the first i preset time periods to obtain the i-th statistical result; i is a positive integer; Perform fitting processing on the i-th statistical result to obtain the distribution parameters of the i-th control deviation; In response to the distribution parameter of the i-th control deviation and the distribution parameter of the i-1-th control deviation not satisfying the preset approximation condition, i is increased by 1, and the steps of performing frequency statistics, fitting processing, and determining the preset approximation condition on the control deviations within the first i preset time periods are repeated; In response to the distribution parameters of the i-th control deviation and the distribution parameters of the i-1-th control deviation satisfying the preset approximation conditions, the distribution parameters of the i-th control deviation are used as the distribution parameters of the control deviation within the reference time period, the time period corresponding to the first i preset time lengths is used as the reference time period, and the end time of the i-th preset time length is used as the starting time of the new reference time period. For the new reference time period, the steps of performing frequency statistics, fitting processing and judging the preset approximation conditions on the control deviations within the first i preset time lengths are repeated.

20. The control system of claim 18, wherein: The second controller is further configured to: Discretizing the estimated distribution parameters of the control signal at the current moment to obtain a plurality of control signals and a control signal probability of each control signal; Discretizing the distribution parameters of the control deviations within the reference period to obtain a plurality of control deviations and a control deviation probability for each control deviation; Performing a superposition correction process on each control signal and each control deviation to obtain a corresponding corrected control signal, and determining the product of the control signal probability of the control signal and the control deviation probability of the control deviation as the correction probability of the corresponding corrected control signal; Fitting processing is performed on the obtained multiple corrected control signals and the corrected probability of each corrected control signal to obtain the corrected distribution parameter of the control signal at the current moment.

21. The control system according to any one of claims 13 to 17, wherein: The control decision maker is further configured to: The control signal at the current moment is determined according to the estimated distribution parameter of the control signal at the current moment, the corresponding relationship between the control signal and the load, and the reference load range.

22. The control system of claim 21, wherein: The control decision maker is further configured to: Determining, based on the estimated distribution parameter at the current moment, the correspondence between the control signal and the load, and the reference load range, a probability that the load falls within the reference load range as a reference probability; In response to the reference probability being greater than or equal to a probability threshold, determining a reference control signal range corresponding to the reference load range; Determining the power generation corresponding to a plurality of reference control signals within the reference control signal range according to the corresponding relationship between the control signal and the power generation; The control signal at the current moment is determined from the multiple reference control signals according to the power generation amounts corresponding to the multiple reference control signals.

23. The control system of claim 22, wherein: The control decision maker is further configured to, in response to the reference probability being less than the probability threshold, determine a reference control signal corresponding to an upper limit value of the reference load range as the control signal at the current moment. 24 . A computer-readable storage medium, wherein when instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor is prompted to execute the control method for a wind turbine generator set according to claim 1 .

25. A computer device comprising: at least one processor; at least one memory storing computer-executable instructions, When the computer-executable instructions are executed by the at least one processor, the computer-executable instructions prompt the at least one processor to execute the control method for the wind turbine generator set according to any one of claims 1 to 11.

Images